Class Cleavage Roots and Left Electoral Mobilization in Western Europe Vincenzo Emanuele

Article

Lost in Translation? Class Cleavage Roots and Left Electoral Mobilization in Western Europe Vincenzo Emanuele

I investigate whether the strength of the class cleavage in Western Europe still “translates” into the electoral mobilization of the left. This research question is addressed through comparative longitudinal analysis in nineteen Western European countries after World War II. In particular, the impact of class cleavage is investigated by disentangling its socio-structural (working-class features) and organizational (corporate and partisan) components, thus accounting for its multidimensional nature. Data show that both components have a signiﬁcant impact in Western Europe after 1945. However, while the socio-structural element is still nowadays a substantial predictor of left electoral mobilization, the impact of the organizational element has decreased over time and has become irrelevant in the last twenty-ﬁve years. Therefore, the class cleavage is not entirely lost in translation, but left electoral mobilization is no longer dependent upon the organizational features of trade unions and political parties that originally emerged to represent working-class interests.

investigate whether and under what conditions the (i.e., the presence of a strong industrial working class, high I electoral mobilization of the Western European left is membership rates in trade unions and left-wing parties) predicted by the strength of the class cleavage. The and the electoral support for class bloc parties starting point of this investigation is represented by Bar- (i.e., communist, socialist, and social democratic parties). tolini’sinfluential work (2000), a macro-historical analysis However, Bartolini’s empirical results are drawn from the of thirteen Western European countries in the period “golden age” of class politics, where the working class was 1860–1980. Bartolini found a robust and enduring asso- strong and mainly industry-based, and class bloc parties ciation between different aspects of the class cleavage and trade unions were powerful mass organizations. More- over, Bartolini’s analysis does not properly take into account the interactive dimension between class cleavage A list of permanent links to Supplemental Materials provided aspects and other factors that may moderate the associ- by the authors precedes the References section. ation between class cleavage roots and left electoral mobilization. Amongst the many factors, the most important Data replication sets are available in Harvard Dataverse at: one is undoubtedly the evolution of this relationship https://doi.org/10.7910/DVN/YRLGNR. across time. After several decades where the class cleavage was considered as the basis of politics and electoral competition and all else was instead “embellishment and Vincenzo Emanuele is Assistant Professor in Political ” Science at Luiss, Rome ([email protected]). He is Co-Chair detail (Pulzer 1967), since the 1980s a large body of of the Research Network on Political Parties, Party Systems literature working on micro-level data has emphasized the and Elections of the Council of Europe. His research, focusing decline of the class cleavage (Flanagan and Dalton 1984; mainly on elections and party system change, has appeared in Franklin, Mackie, and Valen 1992; Nieuwbeerta and Comparative Political Studies, West European Politics, Ultee 1999; Dalton 2002; Knutsen 2018). As a result, Party Politics, South European Society and Politics, and the validity of the link between left electoral mobilization Government and Opposition, among others. His mono- and its historical class cleavage roots has to be put into graph, Cleavages, Institutions, and Competition: Under- question in recent decades, characterized by working-class – shrinking and transformation, trade unions and left par- standing Vote Nationalisation in Western Europe (1965 ’ 2015), has been published by Rowman & Littlefield/ECPR ties organizational decline, and the alleged ideological Press. shift of class bloc parties away from economic left goals

doi:10.1017/S1537592721000943 © The Author(s), 2021. Published by Cambridge University Press on behalf of the American Political Science Association. This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited. 1

(Mair 2008; Best 2011; Dalton 2013; Jansen, Evans, and Class Cleavage and Left Electoral De Graaf 2013; Rennwald and Evans 2014; Evans and Mobilization: Theoretical Background Tilley 2012, 2017; Goldberg 2020). In other words, is the class cleavage still able to translate into the electoral The concept of cleavage lies in path-breaking studies by support for the left? Or, instead, is left parties’ support Lipset and Rokkan (1967) and Rokkan (1970). However, completely detached from its historical class cleavage notwithstanding their extensive use of the concept for the fl formulation of their genetic theory of party systems, Lipset roots? And which other intervening factors in uence this fi relationship? and Rokkan do not provide a clear de nition of cleavage, whose meaning has remained loose and has been used for I address these research questions through a compara- fl tive longitudinal analysis in nineteen Western European a long time as a synonym of division and con ict. As countries in the period 1946–2018. In particular, the Bartolini (2000, 16) puts it, the concept of cleavage is “either reduced down to that of social cleavage or raised up impact of class cleavage strength on the electoral mobil- ” ization of the class left is investigated through ecological to that of political cleavage. This is because the concept data and by disentangling the class cleavage into different refers to both social structure and political order. Thus scholars often use the expression “social cleavage” or components related to its socio-structural and organiza- “ ” tional elements, thus accounting for its multidimensional political cleavage depending on which aspect they want nature properly (Bartolini and Mair 2007). to emphasize. Empirical analyses show that the different components In this regard, an important systematization comes from of the class cleavage have a significant impact on the Bartolini and Mair (2007, 199). According to these electoral support for the class bloc in Western Europe in authors, the concept of cleavage incorporates three elem- – ents: 1) the empirical referent of the concept, that is, the the period 1946 2018. In particular, all else equal, a fl sizeable and industry-based working class and strong social group(s) arising from a con ict generated in the corporate and partisan organizations increase the electoral society, such as the working class (and, in reaction to it, the support for the class bloc. However, the temporal evolu- bourgeoisie) in the case of the class cleavage (empirical or tion of the association between class cleavage roots and left socio-structural element); 2) a set of values and beliefs electoral mobilization shows that, while the socio- providing the social group(s) with a sense of identity and structural element is still nowadays a substantial predictor self-consciousness (cultural or normative element); and 3) of the electoral support for the left, the impact of the an organizational structure (i.e., a political party) that organizational aspect has decreased over time and has coordinates and inspires the collective action of the social fi fi group and brings its interests into the political system become no longer signi cant in the last twenty- ve years. fi Therefore, the class cleavage is not entirely lost in trans- (organizational element). This widely accepted de nition lation, as working-class characteristics are still relevant for resulting in the concept of cleavage involves at the same predicting the electoral success of the left. However, the time socio-structural, ideological, cultural, organizational, latter is no longer dependent upon the organizational and institutional elements. Therefore, it follows that a social conflict per se is not sufficient to have a cleavage. To features of trade unions and political parties that originally fl emerged to represent working-class interests. And even the be translated into a cleavage, a social con ict needs to programmatic emphasis left-wing parties put on economic generate a sense of identity and common beliefs and, most left goals—which, in contrast to many scholarly claims, importantly, it needs to be politicized through the activity has not declined in recent years—is not sufficiently able to of an organization devoted to promoting its interests. At the same time, ideological or political issues without moderate the association between class cleavage roots and fi left electoral mobilization. clearly de ned social groups opposing each other cannot fi 1 The article is organized as follows: in the next section I be de ned as cleavages. critically review the literature on class cleavage and its Out of the four cleavages originally pointed out by electoral consequences in Western European societies; the Lipset and Rokkan, the class cleavage was the last to following section is devoted to raising hypotheses about emerge but soon became the most important and most the impact of the different class cleavage aspects on class enduring one. It stemmed from the Industrial Revolu- bloc electoral support; then I introduce the dependent tion and gave rise at the corporate level to trade unions, variable, discuss the methodological choices underlying and at the political and electoral levels, to socialist and the classification of parties within the class bloc, the social democratic parties that politicized the interest of the rising working class. While the other cleavages national and temporal variations of their electoral support, — and their programmatic emphasis; in the subsequent emerged only in certain contexts and not in others according to specific cultural, socio-economic, or polit- section I operationalize the independent and control — variables and discuss the method used in the empirical ical conditions the class cleavage spread across Europe analysis; lastly, I show the results of the empirical analysis; and was politicized everywhere during the historical a conclusion follows. phases of democratization, enfranchisement, and

2 Perspectives on Politics

Downloaded from https://www.cambridge.org/core. IP address: 170.106.35.93, on 28 Sep 2021 at 03:22:57, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1537592721000943 development of mass politics (Lipset and Rokkan 1967; On the other hand, the decline in class voting is also the Rokkan 1970). result of globalization and de-industrialization processes. Given it was a so-called functional cleavage, with The industrial working class of Western democracies has opposing social groups (working class vs. employers) shrunk in size, thus becoming electorally irrelevant (Best based on ideas and interests and cutting across the 2011). In other words, a “structural dealignment” has territory of the respective country, the class cleavage taken place, according to which “the importance of a historically played the role of a “homogenizing cleavage” cleavage has weakened simply because of a reduction or in the process of nationalization of politics. Indeed, it even ‘extinction’ of a social group, and not because the created nonterritorial nationwide alignments with oppos- linking issue or subject has lost effect on vote choice” ing socialist and social democratic parties representing (Goldberg 2020, 70). the working class and with conservative and liberal Beyond these two bottom-up perspectives, a top-down parties advocating the threatened interests of the bour- approach focusing on party strategies has also emerged. geoisie (Caramani 2004). According to this strand of literature, the weakened link As a fundamental pillar of the well-known “social between social classes and reference political parties is not cleavage model” (Lipset and Rokkan 1967), the alignment symptomatic of a dealignment of class. Rather, it is due between the working class and its reference parties became to top-down processes in which political parties, particu- “frozen” in the 1920s, modeled the shape of Western larly social democratic ones, have moderated their European party systems for decades and contributed to ideological orientation and reduced the emphasis on their long-term stability (Rose and Urwin 1970; Bartolini traditional economic left goals, thus favoring a substan- and Mair 2007; Oskarson 2005). Once structured in tial convergence with the mainstream right (Mair 2008; Western European countries, this cleavage became the Dalton 2013; Evans and Tilley 2012, 2017). Thus, left- foundation of the widespread left-right dimension of wing parties have appealed to different social groups, conflict that crystallized everywhere in Western Europe like public sector employees and socio-cultural profes- as the main dimension of political opposition in each party sionals (Przeworski and Sprague 1986;Kitschelt1994; system (Fuchs and Klingemann 1990; Mair 2007). Gingrich and Häusermann 2015; Benedetto, Hix, and Since the 1980s, the great bulk of studies about the class Mastrorocco 2020). However, in so doing, they risk cleavage turned towards a micro-level perspective focusing alienating the support of their traditional working-class on class voting, namely the association between voters’ base (Karreth, Polk, and Allen 2013;Abou-Chadiand occupational class and vote choice. Although individual Wagner 2019).3 class voting is not the object of this study, it is important to While the top-down approach has the merit of putting underline some key findings of this literature that are party strategies back in the discussion, the role of parties deeply intertwined with the macro-level research about (and trade unions) as organizational agencies of the class the socio-structural and organizational roots of left elect- cleavage is a systematically neglected aspect of the litera- oral mobilization. In this regard, since the 1980s the ture on class voting, which instead focuses mostly on the literature on class voting has observed a decline in the socio-structural element of Bartolini and Mair’s concep- explanatory power of class differences on vote choice tualization (2007) introduced earlier. Indeed, the organ- (Flanagan and Dalton 1984; Franklin, Mackie, and Valen izational element is properly recognized as an important 1992; Nieuwbeerta and Ultee 1999; Dalton 2002; Knut- aspect of the class cleavage in the macro-historical study by sen 2018).2 However, scholars disagree about the leading Bartolini (2000). He focuses on the socio-structural and process causing such dealignment between class and vote the organizational elements of the class cleavage and finds choice. a consistent and fairly stable link between class cleavage On the one hand, a “behavioral dealignment” emerges aspects and left electoral mobilization. Despite the pres- as a result of a “proactive change of the electorate’s ence of relevant national variations, he observes an overall preferences” (Goldberg 2020, 86). A generalized increase robust and enduring association between the presence of a in societal affluence and better working conditions lead strong industrial working class, high membership rates in post-materialist values and “new politics” issues to enter trade unions and left-wing parties, and electoral support the political debate. Consequently, individuals are more for class bloc parties (communist, socialist, and social and more mobilized by individual leaders or according to democratic parties).4 However, the current validity of specific issues (i.e., environment, multiculturalism, Bartolini’s empirical results should be put into question European integration) cutting across traditional class loy- for two main reasons. alties (Inglehart 1984; Kriesi 1998; Dalton 2002; Hooghe, First, the results he gets are drawn from the “golden age” Marks and Wilson 2002; Garzia 2014). As a result of of class politics, where the working class was mainly changing habits and preferences, class divisions have industry-based, and class bloc parties and trade unions become blurred and unable to explain differences in voting were strong mass organizations. By working on a different behavior (Clark and Lipset 2001). time frame (i.e., 1946-2018) characterized by working-

class shrinking and transformation, trade unions and left class and the entire working class and indicates to what parties organizational decline, and the previously discussed extent the dependent labor force (i.e., the potential social ideological shift of class bloc parties, will left electoral base of the class cleavage) is industry-based rather than mobilization still be linked to the class cleavage’s socio- employed in agriculture or services. As for the size, the structural and organizational roots?5 larger the industrial working class, the stronger the Second, Bartolini’s empirical analysis does not properly “social pressure of the group on its members toward take into account the interactive dimension between class attitudinal and behavioral conformity” (Bartolini 2000, cleavage aspects and other important factors. For instance, 145). As a result, working-class size “fosters socialist as the socio-structural element is considered to be political recruitment and mobilization” (ibid., 146). To dependent upon the organizational one for the transform- a similar extent, homogeneity is at least as important as ation of the social conflict into a proper cleavage (Bartolini size. Indeed, regardless of its overall size, Bartolini (2000, 2000, 16-25), does the interaction between two class 152-153) argues that the greater potential for class bloc cleavage elements positively affect left electoral mobiliza- electoral mobilization lies in the industrial sector rather tion? Moreover, as class is not the only source of identity than in agriculture—whose proletariat is organizationally and conflict within a society, does the presence of rooted unstable, volatile, and unable to exploit collective action. cultural (i.e., ethnic, linguistic, religious) cleavages cross- Moreover, service workers were traditionally considered cutting class loyalties undermine the association between either “an essentially conservative element within mod- class cleavage roots and left electoral mobilization? Finally, ern society” (Goldthorpe 1982, 180), or more likely to as emphasized by the top-down approach of class voting have postmaterialist or libertarian values compared to literature, party strategies are expected to influence the industrial workers (Inglehart 1977; Knutsen 2018). relationship between the socio-structural element (i.e., the More recently, however, the literature on class voting working class) and the support for class bloc parties. has tried to account for the increasing heterogeneity in the Therefore, another question to be addressed is whether occupational structure, thus extending the working class the emphasis class bloc parties put on traditional economic beyond the industrial sector: scholars working on the left goals affects the association between the strength of the “EGP class scheme” (Erikson, Goldthorpe, and Portocar- working class and the support for these parties. In the ero 1979) and those relying on the Oesch scheme (Oesch following sections, I try to take a step forward in the study 2006) also consider farm laborers and service workers as of class cleavage roots and their electoral consequences among the working class.8 This more encompassing inter- by empirically addressing all the mentioned research pretation of the working class leads us to consider, besides questions. a measure for those employed in the industrial sector, a measure covering the total working class, thus including Hypotheses the dependent labor force in agriculture and services. Our In the effort of tracing class cleavage roots behind left first hypothesis is as a result: electoral mobilization, we need to start from the socio- HYPOTHESIS 1: The more sizeable and homogeneous the structural and organizational elements of the class cleavage working class, the larger the electoral support for the introduced in the previous section.6 class bloc. The socio-structural element of the class cleavage, its reference social group, is the working class. It is straight- The organizational element of the class cleavage forward to hypothesize that a more favorable environ- involves both its corporate and partisan dimensions. The ment for the electoral success of the class left is a strong presence of strong organizations for the defense of the reference social group, namely a large and cohesive corporate and political interests of the working class is working class. While the latter can be generally con- expected to be a crucial prerequisite for the electoral sidered as the group of people employed for wages,7 the success of the class bloc. The underlying idea is that the sectors considered relevant to forming the socio-struc- more voters are encapsulated into the organization of trade tural element of the class cleavage can be a matter of unions, and even more, political parties representing their debate. The traditional conceptualization and interpret- interests, the easier for such organizations to mobilize their ation of the “owner-worker” cleavage (Lipset and Rok- putative constituency of voters (Bartolini 2000).9 Hence, kan, 1967) referred narrowly to the industrial working the second hypothesis: class. In particular, Bartolini (2000) argued that the HYPOTHESIS 2: The denser the organizational encapsula- strength of the working class could be mainly captured tion, the larger the electoral support for the class bloc. by two indicators, size and homogeneity. The former is the ratio between industrial workers and active popula- Besides these two main hypotheses, we also expect some tion in a country and addresses the question of how large interactive effects involving the two main class cleavage the putative social base of the class cleavage is. The latter, aspects and their relationship with time or other context- instead, refers to the ratio between the industrial working ual factors.

4 Perspectives on Politics

Downloaded from https://www.cambridge.org/core. IP address: 170.106.35.93, on 28 Sep 2021 at 03:22:57, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1537592721000943 To begin with, as the two aspects of the cleavage are the working class, thus potentially undermining its ideo- deeply intertwined, we expect a positive interaction logical cohesion and sense of identity (see Rokkan 1970; between them. On the one hand, the impact of a dense Bartolini and Mair 2007; Lijphart 1999; Nieuwbeerta and organizational encapsulation on the electoral support Ultee 1999). Therefore, cultural (i.e., ethnic, linguistic, or for the class bloc will be stronger the more powerful the religious) heterogeneity can be understood as a factor reference social group is. This is because the organiza- undermining the development of a strong identity among tional side of the cleavage is supposed to represent the the members of the working class. As a result, the positive armor of the otherwise politically defenseless working marginal effect of social group strength and organizational class: following this parallelism, the efficacy of the density on class bloc electoral support is expected to armor will also depend upon the strength of the body decrease as cultural heterogeneity increases. Putting it that wears it. On the other hand, as interaction effects differently, when other sources of membership are always symmetrical (Berry, Golder, and Milton (i.e., ethnicity, language, religion) threaten or challenge 2012), we also expect that the marginal effect of the the socio-structural and organizational aspects of the class socio-structural characteristics of the class cleavage on cleavage, the latter will have a lower impact on the electoral left electoral mobilization will be stronger the denser return of the class bloc. Therefore, our fifth hypothesis: the organizational network of the cleavage is itself. HYPOTHESIS 5: The marginal effect of the socio-structural Indeed, the positive electoral mobilization effects trig- and organizational aspects of the class cleavage on the gered by a sizeable and homogeneous working class electoral support for the class bloc decreases as cultural (i.e., a strong social group) are likely to be further heterogeneity increases. stimulated and multiplied by the presence of rooted corporate and partisan organizations that translate Moreover, we also believe that supply-side strategies working-class interests from the societal to the political play a role in the story (Dalton 2013; Jansen, Evans, and arena. Therefore: De Graaf 2013; Rennwald and Evans 2014; Evans and Tilley 2012, 2017). Indeed, as class bloc parties are HYPOTHESIS 3: The marginal effect of the organizational expected to represent the interests of their reference social aspect of the class cleavage on the electoral support for group (i.e., the working class), the association between the class bloc increases as the socio-structural aspect working-class size and homogeneity and support for class becomes stronger, and vice versa. bloc parties will be closer the more class bloc parties are Regarding the impact of class cleavage aspects across perceived as the legitimate delegates to represent working- time, scholars have long debated about the resilience (Rose class interests. This outcome is expected to occur as far as and Urwin 1970; Bartolini and Mair 2007; Evans 2000; class bloc parties emphasize traditional economic left Elff 2007) vis-à-vis the decline of the class cleavage goals (e.g., welfare expansion, market regulation, con- (Inglehart 1984; Franklin, Mackie, and Valen 1992; Clark trolled economy, equality, support for labor groups), thus and Lipset 2001; Dalton 2002; Knutsen 2018; Goldberg continuously and consistently feeding and reinforcing 2020). Yet most of this literature has addressed this topic working-class loyalty. Conversely, as far as such parties through individual-level data to assess class voting decline. de-emphasize traditional economic left goals in favor of Conversely, I will test the effect of class cleavage roots on other non-class-related policies, they become increasingly left electoral mobilization over time through aggregate- perceived as outsiders and their linkage to the working level data that, “while less informative, offer a necessary class progressively fades (Sartori 1982, 152-153). At the longer-term perspective” (Bartolini and Mair 2007, 62). same time, the impact of the organizational element Consistently with the theory claiming class cleavage (especially CPD) on left electoral mobilization is also decline and dealignment in Western societies, we expect expected to be influenced by the extent to which such left electoral mobilization to be less linked to its class organizations emphasize their reference to class interests. cleavage’s socio-structural and organizational roots as time In other words, if class bloc parties are not perceived as goes by. As a result: credible in their attempt to represent working-class interests, their organizations, albeit strong, become as “empty HYPOTHESIS 4: The marginal effect of the socio-structural vessels” (Katz and Kolodny 1994) as far as left electoral and organizational aspects of the class cleavage on mobilization is concerned. Therefore, we expect the fol- the electoral support for the class bloc decreases lowing: over time.

As far as other contextual factors are concerned, the HYPOTHESIS 6: The marginal eﬀect of the socio- presence of cultural cleavages is certainly expected to structural and organizational aspects of the class interact negatively with class cleavage aspects. Indeed, if cleavage on the electoral support for the class bloc the society is internally divided by ethnicity, language, or increases as class bloc emphasis on traditional left religion, this means that other allegiances also “crosscut” economic goals increases.

fi The Class Bloc: Composition, Electoral ninety-eight class bloc parties resulting from this classi - cation. Support, and Programmatic Emphasis fi fi Once the class bloc is de ned and the parties belonging The empirical analysis starts from the identi cation and to it are identified, the second task is to assess variation in operationalization of the dependent variable, the electoral the class bloc electoral support. The latter can be oper- mobilization of the class left. As we are interested in a ationalized as the aggregate vote share collected by class systemic-level study, where the observation is not an bloc parties in a given country in a given parliamentary individual party politicizing the class cleavage but instead election (Lower House). the overall electoral strength of all parties politicizing it in a Overall, the average electoral support for class bloc given time and space, we will focus on an aggregate parties is 38% in the 349 elections under study. As clearly measure, namely the class bloc electoral support. displayed by the quadratic line in figure 1, the electoral The first task is the definition of the “class bloc” and the fi support of class cleavage has proved to be substantially identi cation of parties belonging to it. In this regard, by stable over time, with just a 3.5 percentage points variation relying on Bartolini and Mair (2007) and Bartolini (2000), over almost sixty-five years, namely from 1946 to 2010 we consider the historical communist, socialist, social (from the peak of 40.4% in the 1970s to 36.9% in the democratic, and labor parties that originally emerged to 2000s). Finally, a noticeable reduction is shown in the last mobilize the working class and to represent its interests in “ decade (31.9% on average). national political life. In other words, we include those If we exclude Southern European countries where parties which are the historical product of the structuring democracy comes back in the 1970s after the fall of the of the working class movement” (Bartolini and Mair 2007, “ ” authoritarian regimes (Cyprus, Greece, Portugal, and 46), while parties of the new politics (Poguntke 1987; Spain), and we focus only on the remaining fifteen coun- Müller-Rommel 1989), which emphasize issues like envir- tries where we have data points since the end of World onmentalism, feminism, or civil rights, have always been War II, the electoral decline of the class bloc is more excluded. pronounced, with a range going from 40.2% in the 1960s From a methodological viewpoint, we have excluded to 27.6% in the 2010s. This means that to some extent the parties that have never received 1% of the vote share in historical electoral decline of the class bloc in Nordic and parliamentary elections. This choice was made to exclude Western European countries has been supplemented by minor parties whose potential inclusion in the class bloc ffi the entry of electorally powerful class blocs in third-wave was hard to assess with a su cient degree of precision countries. At the same time, it highlights the importance given the scarce information available. of looking at national variations to gauge the phenomenon As a result, for thirteen Western European countries under study in greater detail.11 (Austria, Belgium, Denmark, Finland, France, Germany, However, one may argue that the detected resilience of Ireland, Italy, Netherlands, Norway, Sweden, Switzer- class bloc electoral support—albeit with a clear downward land, and United Kingdom) included in the original trend in the last decade—has little or nothing to do with studies by Bartolini and Mair (2007, 285) and Bartolini the hold of the class cleavage. Conversely, one may (2000, 573–574) that cover the period up to the 1980s, we ’ hypothesize that, while the class cleavage has deeply followed these authors criteria. However, as the class declined in Western European societies in terms of social cleavage is not only a historical product but a dynamic composition and organization, parties that once appealed concept, we have also carefully assessed the potential to the working class are still able to maintain a fairly stable inclusion of all those parties that are direct successors of electoral support as they have reoriented their program- traditional working-class parties or new parties emphasiz- matic platform by de-emphasizing economic left issues ing traditional left issues. A qualitative assessment was and promoting a catchall appeal (Kirchheimer 1966; Mair needed to include class bloc parties in the aforementioned 2008; Dalton 2013; Evans and Tilley 2012, 2017). In thirteen countries in recent years and in the remaining six other words, we could be in front of an issue of misplaced countries (Cyprus, Greece, Iceland, Luxembourg, Portu- equivalence (Adcock and Collier 2001), as traditional left gal, and Spain) in the whole period. We have carefully parties have maintained the same name and symbol but considered all parties belonging to the communist, social- have changed their party platform compared to the 1960s ist, and social democratic party families in the ParlGov or 1970s and no longer emphasize traditional class left database (Döring and Manow 2019) and in the commun- goals. Testing empirically whether class bloc parties have ist and socialist party families in the Comparative Mani- shifted ideologically away from traditional economic left festo Project Database (CMP) (Volkens et al. 2019). fi goals over time is, therefore, a crucial preliminary step to Then, by relying on the speci c literature on these parties verify the validity of our operationalization with reference and party-related sources, we have excluded those parties to the class cleavage. Consistently with recent comparative that did not fit our substantive or methodological cri- 10 and longitudinal analyses (Jansen, Evans, and De Graaf teria. Table A1 in the online appendix reports the list of 2013; Adam and Ftergioti 2019) our empirical tests based

6 Perspectives on Politics

Note: The observations in the ﬁgure represent the aggregate vote share collected by class bloc parties in a given parliamentary election (Lower House). Class bloc parties are communist, socialist, social democratic, or labor parties that have received 1% of the vote share in parliamentary elections at least once in their electoral history. The complete list of class bloc parties is reported in table A1 in the online appendix.

on data from the CMP between 1945 and 2018 (Volkens r = .796), we have combined the two variables expressed in et al. 2019) finds unequivocally that class bloc parties have standardized forms into the index of Social group strength. not reduced their emphasis for traditional economic left The level of organizational density can be captured goals over time (refer to table A2 and the related discussion through the use of two indicators (Bartolini and Mair in the online appendix). Therefore, the selected class bloc 2007, 231–38; Bartolini 2000, 262–63). The first refers to parties can be safely considered as the legitimate represen- the corporate channel and is Trade union density (TUD), tatives of the working-class side of the class cleavage. which is the ratio between union membership and the total dependent labor force of a country. TUD data have Data and Method been collected from Visser (2016).13 The second indicator The operationalization of our main independent variables instead refers to the partisan channel and is the ratio follows. Data about Industrial working-class size, Total between class bloc parties’ membership and the total working-class size, and Working-class homogeneity have been electorate (Class partisan density, CPD). As regards CPD, taken from Bartolini (2000) for earlier years (up to the the source for the collection of data on class parties’ 1970s), while for later ones the source is the International membership has been the MAPP Project Data Archive Labour Organization (ILO), supplemented by data from (Van Haute and Paulis 2016).14 As in the case of socio- the Organization for Economic Cooperation and Devel- structural factors, also TUD and CPD are tested separately opment (OECD) in case of gaps in the ILO data.12 and then combined into a standardized index of Organ- Following international standards and previous literature izational density. (e.g., Mainwaring and Zoco 2007), we define industrial By looking at figures 2 and 3 we get a synthetic scheme workers as those dependent laborers employed in mining, of the temporal evolution and national variations of the manufacturing, construction, and transport. Preliminary socio-structural and organization aspects of the class cleav- regression models, reported in the online appendix (table age. To begin with, consistently with the vast literature A4), show that it is precisely the size of the industrial emphasizing working-class shrinking and left parties’ working class, rather than that of the working class as a organizational decline (Dalton 2002, 2013; Mair 2008; whole (including agriculture and service workers) that Best 2011; Goldberg 2020), our data show that all class- matter for left electoral mobilization. Consequently, we cleavage related variables decline over time.15 This general have excluded total working-class size from the following picture is fairly consistent across countries, as displayed by analyses and, given that industrial working-class size and figure 2, which contrasts the temporal evolution of the two working-class homogeneity are highly collinear (Pearson’s standardized indices of social group strength and

Figure 2 Temporal evolution of social group strength and organizational density (1946–2018)

Note: Social group strength and organizational density are synthetic indices combining variables expressed in standardized forms. The former combines industrial working-class size and working-class homogeneity. The latter combines class partisan density and trade union density. Both indices have been rescaled to a 0–100 range.

organizational density (both rescaled to a 0–100 range to other factors (H4, H5, and H6). Specifically, Time is facilitate the comparison). Moreover, by contrasting coun- operationalized as the number of years elapsed since try averages, as in the scatterplot of figure 3, interesting 1945. The addition of a trend variable is also particularly insights emerge. Indeed, the association between the two suitable in time-series analysis, as it provides control to class cleavage-related aspects is lower than one might problems of spurious correlations (Roberts and Wibbels expect (r = .215). This means that the development of 1999). The level of Cultural heterogeneity is assessed dense organizational networks (i.e., strong trade unions through the Fractionalization index (Alesina et al. 2003). and left parties) is not necessarily linked to the presence of The index ranges from 0 to 1 where 0 means perfect a sizeable and industry-based working class. This is the homogeneity (the whole population in a country has the case of Germany and Switzerland—where a strong social same ethnicity, language, and religion), and 1 means group goes hand in hand with a comparatively weak perfect heterogeneity (there is a 100% probability that organizational density—but also of Cyprus and Iceland, two randomly selected individuals within a population displaying the opposite situation of relatively strong organ- belong to different ethnic, linguistic, or religious groups). izations in a context of a small and heterogeneous working As Alesina et al. (2003) do not provide cross-temporal class. Not surprisingly, Scandinavian countries like Swe- variation, we have collected data from Bartolini (2000) den and Denmark as well as Austria show the strongest until the 1970s, and Patsiurko, Campbell, and Hall class cleavage roots, while Greece and Ireland fall at the (2012) for the most recent period.17 To measure Class bottom of the rank.16 bloc emphasis on traditional economic left goals we have The impact of socio-structural and organizational roots selected the following items in the CMP (Volkens et al. of the class cleavage on class bloc electoral support is 2019): market regulation (per403), economic planning investigated also by looking at interaction effects with (per404), control of the economy (per412), Marxist

8 Perspectives on Politics

analysis (per415), equality (per503), welfare state units (countries). In this context, problems of heterosce- (per504), and labor groups (per701).18 Then we have dasticity and autocorrelation may arise (Stimson 1985). built an index that weighs, within each bloc, the individual Diagnostic tests such as an LR-test for panel- party emphasis for the respective party’s electoral support. heteroskedasticity (p<.001) and a Wooldridge test of The empirical analyses will also be integrated by some autocorrelation (p<.05) (Drukker 2003) confirmed such other control variables that may have effects on the concerns. Moreover, we have performed a Hausman test dependent variable. To begin, we control for the level of to compare the results of fixed and random effects models. class bloc “Organizational cohesion” (Bartolini 2000), The result is that H0 (i.e., the difference in the coefficients namely the number of parties included in the class bloc is not systematic) is not rejected and hence we can use a in a given country in a given election.19 We will then random effects model. Moreover, given that some covari- control for systemic properties such as the overall level of ates in the models are almost or completely time-invariant ideological polarization in the country (through Dalton’s (e.g., average district magnitude, the dummy for third- Polarization index (2008)20 and the electoral system wave democracies), the inclusion of country-fixed effects is (through the natural log of the Average district magnitude, not recommendable (Beck and Katz 2007; Mainwaring, which is calculated by dividing the total number of seats to Gervasoni, and España-Najera 2017; Imai and Kim be allocated by the total number of districts).21 Further- 2019). more, we will also add a dichotomous variable for Third- Hence, in the next section, the previously raised wave democracies (Huntington 1991), namely Cyprus, hypotheses will be tested through a model specification Greece, Portugal, and Spain, where democracy came back that is able to tackle such issues deriving from the data only in the 1970s, and the class cleavage emerged in structure. We opted for a Prais-Winsten regression, different structural conditions compared to the rest of namely a panel-corrected standard errors method Western Europe. (PCSE) with a first-order autoregressive parameter Figure 4 reports a synthetic scheme of all hypotheses, (AR1).22 factors, measures, and expected effect on class bloc vote share, while table A3 in the online appendix reports the Empirical Analysis descriptive statistics of the dependent and (unlogged) Table 1 presents the results of the empirical models. independent variables. Model 1 displays the general plain model, while From a methodological viewpoint, we deal with pooled Models 2–8 test the previously hypothesized interaction time-series cross-section data (Beck and Katz 1995) with effects.23 The general plain model tests H1 and H2, repeated observations over time (elections) on the same namely the hypothesized positive effect of, respectively,

Figure 4 Synthetic scheme of hypotheses, factors, measures, and expected effect on class bloc vote share

Expected Effect on Hypothesis Factor Measures Class Bloc Vote Share

Industrial working-class size and Working- H1 Social group strength + class homogeneity Trade union density (TUD) and Class H2 Organizational density + partisan density (CPD) Class cleavage aspects' H3 Social group strength*Organizational density + mutual reinforcement Social group strength/Organizational H4 Class Cleavage and time - density*Time H5 Class cleavage and Social group strength/Organizational - cultural cleavages density*Cultural fractionalization H6 Class cleavage and class Social group strength/Organizational + bloc parties' emphasis density*Economic left emphasis

Controls Cultural heterogeneity Fractionalization index Weighted index of Class bloc emphasis on Economic left emphasis traditional economic left goals Organizational cohesion Number of parties included in the Class bloc Ideological polarization Polarization index Electoral system Average district magnitude (ADM) Dummy variable for Cyprus, Greece, Third-wave democracies Portugal, and Spain Time Number of years since 1945

the socio-structural and the organization elements of the analysis of the interactive models (H3–H6), reported in class cleavage on left electoral mobilization.24 table 1 (Models 2–8). The general plain model (Model 1) shows that both Starting from H3, Model 2 shows that the interaction class cleavage factors emerge as positive and significant between social group strength and organizational density is predictors of left electoral mobilization. As a result, H1 statistically significant (at p<.05), as expected. This means and H2 are confirmed.25 By comparing the effects of social that the marginal effect of organizational density on class group strength and organizational density, the latter shows bloc vote share increases as far as social group strength a comparatively larger impact on left electoral mobiliza- increases, and vice versa. As shown by the marginal plots tion. Indeed, as organizational density increases by one reported in Figure 5, on the one side (right pane), the percentage point, class bloc vote share increases by 0.14 advantage of having a strong organizational network points against 0.11 in the case of a one percentage point becomes tangible only at a certain level of industrial increase of social group strength.26 working-class size and homogeneity. In other words, with Table A5 in the online appendix offers some robustness a meager and heterogeneous (i.e., not industry-based) checks with a replication of Model 1 in Table 1 across working class, a powerful organization is not able to yield three different model specifications: a feasible generalized noticeable electoral returns. On the other side (left pane), least squares (FGLS) regression (Parks 1967) with a the strength of the reference social group has an impact on specification for panel heteroscedasticity and an autore- the electoral support for the class bloc only if there is a gressive parameter (AR1); a panel-corrected standard certain level of organizational encapsulation. errors regression (Beck and Katz 1995) without the spe- Regarding H5, Models 5 and 6 of table 1 yield different cification for autocorrelation; a generalized linear model results. While confirming the hypothesized negative inter- (GLM) estimator with logit link function (Costalli and action between class cleavage aspects and cultural hetero- Ruggeri 2019). Substantive results are consistent with geneity, only social group strength is influenced by those in table 1 across different specifications.27 cultural cleavages in a statistically significant way (the So far, we have tested our first two hypotheses involving interaction term is significant at p<.01). More specifically, the linear effect of the two aspects of the class cleavage on social group strength plays a significant role in the electoral the electoral success for the class bloc. We now turn to the success of the class bloc only when cultural heterogeneity is

10 Perspectives on Politics

Table 1 Determinants of class bloc vote share in Western Europe after 1945 https://www.cambridge.org/core Plain model Interactions models

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 b pcse b pcse b pcse b pcse b pcse b pcse b pcse b pcse . https://doi.org/10.1017/S1537592721000943 Social group 0.111* 0.043 0.024 0.061 0.122† 0.069 0.122** 0.043 0.266*** 0.077 0.111** 0.043 0.037 0.061 0.109* 0.043 strength Organizational 0.137*** 0.038 −0.052 0.095 0.137*** 0.038 0.298*** 0.067 0.121** 0.038 0.182** 0.064 0.139*** 0.038 0.120* 0.055

. IPaddress: density Social group str. ——0.004* 0.002 —— ——— — —————— *Org. density Social group str. ————−0.0003 0.002 ——— — —————— 170.106.35.93 *Time since 1945 Org. density*Time ——————−0.004** 0.002 — — —————— since 1945 , on Social group str. —————— ——−0.560** 0.202 ——————

28 Sep2021 at03:22:57 *Fract. index Org. —————— ——— —−0.249 0.253 ———— density*Fract. index Social group str. —————— ——— — ——0.003† 0.001 —— *Economic left emphasis index , subjectto theCambridgeCore termsofuse,available at Org. density* —————— ——— — ————0.001 0.002 Economic left emphasis index Fractionalization −9.762† 5.622 −9.731† 5.358 −9.929† 5.627 −9.979† 5.494 14.37 10.870 −3.668 8.905 −9.446† 5.657 −9.532† 5.728 index Economic left −0.062† 0.036 −0.068† 0.037 −0.063† 0.037 −0.061† 0.036 −0.065† 0.036 −0.065† 0.037 −0.188* 0.089 −0.081 0.065 emphasis index N. parties in the 1.032† 0.579 1.018† 0.564 1.037† 0.579 0.954† 0.570 0.974† 0.571 1.06† 0.576 1.023† 0.576 1.031† 0.581 Class bloc Polarization index 2.099 4.341 1.711 4.446 2.112 4.346 1.654 4.274 1.984 4.345 2.088 4.379 1.882 4.293 2.114 4.323 Average district 0.79 0.491 0.703 0.481 0.787 0.491 0.888† 0.495 0.728 0.480 0.753 0.489 0.771 0.496 0.809 0.493 magnitude (ln) Third-wave 15.23*** 2.542 14.81*** 2.312 15.22*** 2.536 14.81*** 2.407 15.33*** 2.488 15.34*** 2.470 14.99*** 2.513 15.19*** 2.577 democracies Time since 1945 −0.082† 0.044 −0.073† 0.043 −0.071 0.078 0.046 0.067 −0.086* 0.043 −0.081† 0.044 −0.078† 0.044 −0.084† 0.044 Constant 29.03*** 4.261 33.14*** 4.776 28.61*** 4.945 23.53*** 4.826 23.35*** 4.947 27.72*** 4.611 32.61*** 4.762 29.63*** 4.493 R2 0.479 0.487 0.479 0.491 0.489 0.481 0.483 0.479 Wald χ2 108.07*** 141.83*** 110.98*** 130.62*** 118.87*** 112.12*** 112.63*** 106.11*** N of elections 345 345 345 345 345 345 345 345 N of countries 19 19 19 19 19 19 19 19 11 Note: Prais-Winsten AR1 regressions; panel-corrected standard errors are reported. †p < .10; *p < .05, **p < .01, ***p < .001. Article | Lost in Translation?

Figure 5 Marginal effects on class bloc vote share of the interactions among the two aspects of the class cleavage

below a certain threshold, say 0.34 (80% of the observa- Finally, the results shown in Models 3 and 4 are tions fall below this threshold; see also the marginal plot in surprising as they are partially against the expectation figure 6). Conversely, dense corporate and partisan organ- formulated in H4. Instead of showing unequivocally the izational networks play a positive and significant role on declined class cleavage roots behind left electoral mobil- left electoral mobilization regardless of the extent to which ization in Western Europe, Models 3 and 4 highlight a the society is ethnically, linguistically, or religiously div- fundamental difference between the two aspects of the ided.28 class cleavage. Indeed, as displayed by the steep negative Turning to the interaction effect between the two class slope of the marginal effect plotted in figure 7, the cleavage aspects and class bloc emphasis on traditional association between organizational density and class bloc economic left goals (H6), Models 7 and 8 do not support vote share has dramatically collapsed in recent years. On our expectation. The organizational element of the class the contrary, in the case of social group strength (Model cleavage is shown to be unrelated to the extent that 3), the interaction with time is not significant, suggesting political parties appeal to their putative reference social that the association between working-class features and group. In other words, organizations matter, regardless of class bloc vote share has not substantively declined in the policy proposals they emphasize.29 The interaction recent years. between social group strength and class bloc emphasis on As further robustness checks of H4, we have run separ- traditional economic left goals is positive and significant, ate regressions of the general model (Model 1 in table 1) for but only at p < .10. As a result, the more class bloc parties two time periods, with the cutoff point set in 1989 (see appeal to their putative social group, the closer the linkage table A6 in the online appendix). The first period (1946– between the working class and left electoral mobilization. 1989) shows that the two aspects of the class cleavage are However, following Berry, Golder, and Milton (2012, significant predictors of class bloc electoral support (both at 660-662), that kind of effect is of trivial magnitude and, p<0.001). Conversely, in the second period (1990–2018), therefore, not substantively significant. The correct inter- consistent with figure 7,31 the impact of organizational pretation is therefore that the effect of social group density becomes not significant, while that of social group strength on class bloc electoral support depends only strength maintains a significant impact on left electoral slightly upon the extent to which class bloc parties empha- mobilization (albeit only at p<0.05). As a result, Hp4 is size traditional economic left goals.30 only partially confirmed, and this finding sheds new light

12 Perspectives on Politics

Figure 7 Marginal effect of organizational density on class bloc vote share at different levels of time

on the relationship between left electoral mobilization and In other words, in the last twenty-five years, the elect- its cleavage roots. In contrast to Bartolini’s findings (2000), oral support for class bloc parties has been completely the organizational density, once the most important pre- detached from the organizational density of the corporate dictor of left electoral mobilization, no longer plays a and partisan channels that were alleged to represent significant role in the story.32 working-class interests. However, the historical link

between the industrial working class and the electoral goes beyond Bartolini’s findings by delving deeper into support for class bloc parties has persisted over time and the relationship between class cleavage aspects and can be still detected nowadays. Therefore, notwithstand- other intervening factors. Here, other important ori- ing the scholarly emphasis on class cleavage decline, there ginal findings emerge. are still socio-structural class cleavage roots behind left First, the two aspects of the class cleavage tend to electoral mobilization, although such mobilization no reinforce each other through positive interaction longer goes through the corporate and partisan organiza- effects. Second, the impact of the socio-structural roots tions of the class cleavage. of the class cleavage is also negatively influenced by the level of cultural heterogeneity in the society. Cultural Conclusion divides crosscut the working class and undermine class I have investigated, through comparative longitudinal loyalties. Therefore, as far as ethnic, linguistic, and research design, the association between class cleavage’s religious heterogeneity increases, the effect of the socio- socio-structural and organizational roots and left electoral structural element on left electoral mobilization mobilization in nineteen Western European countries decreases. after 1945. I have advanced the research on class cleavage Third, in contrast to many scholarly claims, class bloc and its electoral consequences in different respects. parties have not reduced their emphasis on traditional First, I have provided an original classification of class economic left goals. However, this supply-side factor bloc parties in Western Europe after 1945. Such classi- seems to be substantially irrelevant in the association fication may travel beyond the scope of this study and between the two class cleavage-related aspects and left become an important contribution for all scholars deal- electoral mobilization. ing with left-wing parties in Europe. Moreover, I have Finally, the temporal evolution of the relationship shown that class bloc electoral support has experienced a between class cleavage roots and left electoral mobilization slight decline over time but is still electorally relevant as is only partially in line with the scholarly claim about class it has gathered about one-third of valid votes in the cleavage decline. Indeed, while the impact of the organ- 2010s. In terms of national variations, the electoral izational aspect has decreased over time and has become no decline is more pronounced in Nordic and Western longer significant in the last twenty-five years, the socio- countries, while Southern European ones have recently structural element has not, and it is still a substantial emerged as those with the largest electoral support for predictor of left electoral mobilization. the class bloc. In a nutshell, the class cleavage is not entirely lost in Most importantly, through an original collection of a translation, as a sizeable and industry-based working class large amount of data coming from multiple sources, I have is still an important predictor today of left electoral operationalized two aspects of the class cleavage, namely its mobilization. But such mobilization is no longer mediated socio-structural and organizational components. By doing by the corporate and partisan organizations, the original so, it is possible to test empirically the impact of these vectors of cleavage translation. Indeed, membership in aspects of class cleavage on its electoral translation, namely trade unions and left-wing parties are no longer associated the electoral support for the class bloc. Moreover, this with left electoral support, but even the extent to which study has also tested how these aspects interact among each left-wing parties emphasize economic left goals is not other as well as with other factors such as the presence of sufficiently able to moderate the association between class cultural cleavages and the programmatic emphasis class cleavage roots and left electoral mobilization. bloc parties put on traditional economic left goals. Finally, These results deserve careful future consideration and the study has also tested how the explanatory power of the have important implications for the study of cleavages and two class cleavage-related aspects vary over time. elections. In particular, they raise fundamental questions The empirical analyses yield relevant results for our about the future of class bloc parties and their class understanding of the link between class cleavage roots cleavage roots. Will the persistent link between class and left electoral mobilization in Western Europe. cleavage’s socio-structural roots and left electoral mobil- Overall, both aspects of the class cleavage have a ization be a sufficient factor for class bloc parties’ electoral significant impact on the electoral support for the class resilience, despite the missing link with class cleavage’s bloc in Western Europe in the period 1946–2018. All organizational roots? Or, instead, will the breakage of the else equal, a sizeable and industry-based working class transmission belt between organizational density and left (but not the agriculture- and service-based working electoral mobilization cause an electoral disintegration of class) and a dense organizational network increase the these parties in the near future? Further research should electoral support for the class bloc. Therefore, our study carefully address these questions, whose answer is of confirms that Bartolini’s findings (2000) still hold in paramount importance for our understanding of class the aggregate also by expanding both the spatial and the politics and electoral competition in twenty-first century temporal scope of its original analysis. But this study Europe.

14 Perspectives on Politics

Downloaded from https://www.cambridge.org/core. IP address: 170.106.35.93, on 28 Sep 2021 at 03:22:57, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1537592721000943 Acknowledgement jobs, which are those where the incumbents hold employment contracts which give them a basic A previous version of this article was presented at the remuneration not directly dependent upon the rev- XXXIII SISP Congress, Lecce, September 11-14, 2019, enue of the unit for which they work.” and at the XII International SISE Conference, Rome, 8 In the EGP class scheme, the categories of manual December 5-6, 2019. I warmly thank Lorenzo De Sio, supervisors, skilled worker, unskilled worker, and farm Alessandro Chiaramonte, Paolo Bellucci, Paolo Barisione, labor are generally considered as part of the working Giorgio Malet, Julia Schulte-Cloos, Elisa Volpi, Davide class; Leiulsfrud, Bison, and Jensberg 2005. In the Angelucci, Alberto Stefanelli, Federico Trastulli, and Oesch eight-class scheme, production workers Bruno Marino for their comments and suggestions on (including employees in the industrial and agricultural earlier versions of the article. sector), service workers, and office clerks, although relying on a different work logic, could be both Supplementary Materials considered as part of the working class; Oesch 2006. To view supplementary material for this article, please visit In more recent works, clerks are excluded from the http://dx.doi.org/10.1017/S1537592721000943. working class; Rennwald and Evans 2014; Oesch and Rennwald 2018. Notes 9 In their recent contribution, Abou-Chadi and Wagner 1 On this point, see the discussion in Bartolini 2000, 2019 show that trade unions still play a crucial role in 16-25. the mobilization of the working class. 2 On the different generations of class voting research, 10 Refer to the online appendix for further discussion on see Knutsen 2007. these criteria. 3 Notice that the three approaches, rather than being 11 In this regard, see figures A1 and A2 in the online independent from each other, are clearly interrelated; a appendix. structural dealignment and the dramatic decline in the 12 The association between ILO and OECD data is size of the traditional class gardée lead left-wing parties almost perfect (Pearson’s r = .92 and .95 for size and to appeal to different social groups in order to suc- homogeneity, respectively). To deal with missing data cessfully compete in the electoral arena. As a result of for the two variables, we have proceeded with linear this programmatic and ideological change, a behav- interpolation. This procedure is logically acceptable ioral dealignment is also an expected outcome, as since the observations in a given country tend to follow social group divisions are less likely to produce clear- a predictable linear trend over time so that each cut vote choices. observation is closely linked to the previous and the 4 Other scholars aiming at measuring class cleavage following ones. strength through ecological data chose to focus on 13 Data refer to net union membership (students, either the organizational aspect only (i.e., trade union unemployed or retired members are excluded). and left parties’ membership density; see Bartolini and 14 For Greece, excluded from the MAPP project, we have Mair 2007; Roberts and Wibbels 1999; Emanuele relied on Mair and Van Biezen 2001 and Van Biezen, 2018) or on both socio-structural (i.e., industrial Mair, and Poguntke 2012. For Luxembourg, the working-class size) and organizational elements; source is Schlager and Weisblatt 2006. Mainwaring and Zoco 2007. 15 Such decline over time is strong and significant for 5 In this regard, Benedetto, Hix, and Mastrorocco 2020 industrial working-class size, working-class homo- find mixed evidence about the relationship between geneity, and class partisan density. By contrast, the share of workers employed in industry and social in the case of total working class and trade union democratic vote share. Indeed, the share of industrial density the negative trend over time is almost workers is a significant predictor of the support for negligible. social democratic parties since 1975, but such associ- 16 For more detailed cross-country variations of all class- ation is no longer significant since 2000. related variables, see figures A3–A9 in the online 6 Consistent with previous literature, the normative appendix. element (i.e., class identity and consciousness) is 17 For Greece, Portugal, and Spain, excluded from Bar- excluded from the analysis as it is not systematically tolini’s study, data until the 1970s have been taken available for the temporal and spatial scope of this from Lane and Ersson 1987, 58, 67. study. On the difficulty to measure this aspect of the 18 With reference to the welfare state and labor variables, class cleavage, see also Bartolini and Mair 2007, we have calculated the difference between the item 208, and Bartolini 2000, 28. “welfare expansion” (per504) and the item “welfare 7 The International Labour Organization (ILO) defines limitation” (per505), and “labor group positive” these as “all those workers who hold paid employment (per701) and “labor group negative” (per702),

respectively. All the other selected items refer to opposition in the legislature preceding the election, positive references and have no corresponding nega- 1 if at least one class bloc party was in government, and tive item. For further details on the index, see the 2 if at least one class bloc party held the position of discussion in the online appendix. prime minister in the same period. The findings do 19 For a brief discussion on this variable, see the online not change by excluding one control variable at a time. appendix. Results are available upon request. 20 Polarization index = 26 All class cleavage-related variables in tables 1 and A4 ( ) – 2 have been standardized and rescaled to a 0 100 range, P ½party L score−party system average L score √ n ðÞ∗ R R ffi i¼1 party vote share 5 , so the respective coe cients can be compared. In terms of standard deviations, if social group strength and organizational density increase by one standard where i represents individual parties; Dalton 2008. deviation, class bloc vote share increase by 2.3 and 3.3 Data on Left/Right positioning for each party are percentage points, respectively. Furthermore, among taken from the CMP; Volkens et al. 2019. As clearly the control variables, the only significant predictor at stated in the literature, high levels of ideological p<0.05 in the general model is the dummy for third- polarization make class issues clearer and party choices wave democracies, which has a powerful positive consequential compared to the opposite situation effect. This means that all else equal, Southern where the ideological differences between parties are European countries have a stronger class bloc com- blurred; Jansen, Evans, and De Graaf 2013; Evans and pared to the rest of Western Europe. Finally, consist- Tilley 2017. ent with Benedetto, Hix, and Mastrorocco 2020, 21 The reference for data collection is Bormann and 5, the economic left emphasis index is negatively Golder 2013. For mixed electoral systems, the total associated with class bloc electoral support, although number of seats is divided by the sum of the number of the relationship is significant only at p < .10. This districts in which seats are allocated in each tier; means that class bloc parties gain votes as far as they Johnson and Wallack 2012. de-emphasize traditional economic left goals. 22 Given that panels are unbalanced (the number of elect- 27 We have also replicated the regressions in table 1 only oral periods within panels ranges between nine and for the period 1946–1980 in the subset of thirteen twenty-six and electoral periods are non- countries (Austria, Belgium, Denmark, Finland, contemporaneous), contemporaneous correlation across France, Germany, Ireland, Italy, Netherlands, Norway, panels (one of the assumptions of the PCSE) has been Sweden, Switzerland, and United Kingdom) included ruled out. Moreover, the Prais-Winsten transformation in Bartolini’s 2000 analysis. Substantive results are the has been preferred to the use of the lagged dependent same as in table 1, with the obvious difference that class variable to preserve sample size and degrees of freedom; cleavage-related aspects do not decline over time until Harbers 2010. Further robustness checks are presented 1980 (i.e., H4, found later in the text). Results are in tables A5–A9 in the online appendix. reported in table A8 in the online appendix. 23 All models are significant (as witnessed by the Wald 28 Although the interaction term is not significant, the Chi-square statistics) and explain a substantive portion coefficient of organizational density is statistically of class bloc vote share in post-World War II Western significant in both Models 1 and 6. Europe (the R-squared ranges between 47.9% and 29 Notice that by replacing organizational density with 49.1%). Moreover, data show no issues of multicol- class partisan density, the substantive result is exactly linearity, as the Variance Inflation Factor in Model the same. 1oftable 1 ranges between 1.2 and 2.4. 30 This result is consistent with Jansen, Evans, and De 24 Moreover, table A4 separately tested the effects of the Graaf 2013, 391, who conclude their analysis by different socio-structural and organizational variables stating that “we did not find evidence for the idea that on left electoral mobilization (see the related discus- left–right positions of left-wing parties alone influence sion in the online appendix). the association between class and vote.” 25 Note also that by adding further controls for party 31 Figure 7 clearly shows that the association between system fragmentation, various measures of turnout, organizational density and class bloc vote share is no economic indicators (GDP growth rate and longer significant since the mid-1990s. unemployment) such results hold. The same applies 32 We have also replicated Models 3 and 4 of table 1 by by controlling for government participation of class replacing the linear time variable with a categorical bloc parties in the previous legislature; see Abou- variable for decades (with the 1950s as the reference Chadi and Wagner 2019; Benedetto, Hix, and Mas- category, see table A7 in the online appendix). The trorocco 2020. Further, we built a categorical variable results confirm the partial rejection of H4, as the having value 0 if all class bloc parties were in the marginal effect of social group strength on class bloc

16 Perspectives on Politics

Downloaded from https://www.cambridge.org/core. IP address: 170.106.35.93, on 28 Sep 2021 at 03:22:57, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1537592721000943 electoral support does not significantly vary over time, Clark, Terry Nichols, and Seymour Martin Lipset, eds. while that of organizational density significantly 2001. The Breakdown of Class Politics: A Debate on Post- declines in the 2010s (see also figure A10 in the online Industrial Stratification. Baltimore: Johns Hopkins appendix). University Press. Costalli, Stefano, and Andrea Ruggeri. 2019. “The Long- References Term Electoral Legacies of Civil War in Young Abou-Chadi, Tarik, and Markus Wagner. 2019. “The Democracies: Italy, 1946–1968.” Comparative Political Electoral Appeal of Party Strategies in Postindustrial Studies 52(6): 927–61. Societies: When Can the Mainstream Left Succeed?” Dalton, Russell J. 2002. “Political Cleavages, Issues, and Journal of Politics 81(4): 1405–19. Electoral Change.” In Comparing Democracy 2: New Adam, Antonis, and Stamatia Ftergioti. 2019. “Neighbors Challenges in the Study of Elections and Voting, and Friends: How Do European Political Parties ed. Lawrence LeDuc, Richard G. Niemi, and Pippa Respond to Globalization?” European Journal of Norris, 319–342. London: Sage. Political Economy 59: 369–84. ——. 2008. “The Quantity and the Quality of Party Adcock, Robert, and David Collier. 2001. “Measurement Systems: Party System Polarization, Its Measurement, Validity: A Shared Standard for Qualitative and and Its Consequences.” Comparative Political Studies Quantitative Research.” American Political Science 41(7): 899–920. Review 95(3): 529–46. ——. 2013. Citizen Politics: Public Opinion and Political Alesina, Alberto, Arnaud Devleeschauwer, William Parties in Advanced Industrial Democracies. Thousand Easterly, Sergio Kurlat, and Romain Wacziarg. 2003. Oaks, CA: CQ Press. “Fractionalization.” Journal of Economic Growth 8(2): Döring, Holger, and Philip Manow. 2019. “Parliaments 155–94. and Governments Database (ParlGov): Information on Bartolini, Stefano. 2000. The Political Mobilization of the Parties, Elections and Cabinets in Modern European Left, 1860–1980: The Class Cleavage. Democracies.” Development version. (http:// Cambridge: Cambridge University Press. www.parlgov.org/). Bartolini, Stefano, and Peter Mair. 2007[1990]. Identity, Drukker, David M. 2003. “Testing for Serial Correlation Competition, and Electoral Availability: The Stabilisation in Linear Panel-Data Models.” Stata Journal 3(2): of European Electorates, 1885–1985. Colchester: ECPR 168–77. Press. Elff, Martin. 2007. “Social Structure and Electoral Beck, Nathaniel, and Jonathan N. Katz. 1995. “What to Behavior in Comparative Perspective: The Decline of Do (and Not to Do) with Time-Series-Cross-Section Social Cleavages in Western Europe Revisited.” Data.” American Political Science Review 89(3): Perspectives on Politics 5(2): 277–94. 634–47. Emanuele, Vincenzo. 2018. Cleavages, Institutions, and Beck, Nathaniel, and Jonathan N. Katz. 2007. “Random Competition. Understanding Vote Nationalization in Coefficient Models for Time-Series Cross-Section Western Europe (1965–2015). London: Rowman and Data: Monte Carlo Experiments.” Political Analysis 15: Littlefield/ECPR Press. 182–95. Erikson, Robert, John H. Goldthorpe, and Lucienne Benedetto, Giacomo, Simon Hix, and Nicola Portocarero. 1979. “Intergenerational Class Mobility in Mastrorocco. 2020. “The Rise and Fall of Social Three Western European Societies: England, France Democracy, 1918–2017.” American Political Science and Sweden.” British Journal of Sociology 30(4): Review 114(3): 928–39. https://doi.org/10.1017/ 415–41. S0003055420000234. Evans, Geoffrey. 2000. “The continued significance of Berry, William D., Matt Golder, and Daniel Milton. class voting.” Annual Review of Political Science 3(1): 2012. “Improving Tests of Theories Positing 401–417. Interaction.” Journal of Politics 74(3): 653–71. Evans, Geoffrey, and James Tilley. 2012. “How Parties Best, Robin E. 2011. “The Declining Electoral Relevance Shape Class Politics: Explaining the Decline of the Class of Traditional Cleavage Groups.” European Political Basis of Party Support.” British Journal of Political Science Review 3(2): 279. Science 42(1): 137–61. Bormann, Nils-Christian, and Matt Golder. 2013. Evans, Geoffrey, and James Tilley. 2017. The New Politics “Democratic Electoral Systems around the World, of Class: The Political Exclusion of the British Working 1946–2011.” Electoral Studies 32(2): 360–69. Class. Oxford: Oxford University Press. Caramani, Daniele. 2004. The Nationalization of Politics: Flanagan, Scott C., and Russell J. Dalton. 1984. “Parties The Formation of National Electorates and Party Systems under Stress: Realignment and Dealignment in in Western Europe. Cambridge: Cambridge University Advanced Industrial Societies.” West European Politics Press. 7(1): 7–23.

Franklin, Mark N., Thomas T. Mackie, and Henry Valen, 7, 2020 (https://dataverse.harvard.edu/dataset.xhtml? eds. 1992. Electoral Change: Responses to Evolving Social persistentId=hdl:1902.1/17901). and Attitudinal Structures in Western Countries. Karreth, Johannes, Jonathan T. Polk, and Christopher S. Cambridge: Cambridge University Press. Allen. 2013. “Catchall or Catch and Release? The Fuchs, Dieter, and Hans-Dieter Klingemann. 1990. “The Electoral Consequences of Social Democratic Parties’ Left-Right Schema.” In Continuities in Political Action: March to the Middle in Western Europe.” Comparative A Longitudinal Study of Political Orientations in Three Political Studies 46(7): 791–822. Western Democracies, ed. M. Kent Jennings and Jan W. Katz, Richard S., and Robin Kolodny. 1994. “Party Van Deth, 203–234. Berlin: De Gruyter. Organization as an Empty Vessel: Parties in American Garzia, Diego. 2014. Personalization of Politics and Politics.” In How Parties Organize: Change and Electoral Change. Basingstoke: Palgrave Macmillan. Adaptation in Party Organizations in Western Gingrich, Jane, and Silja Häusermann. 2015. “The Democracies, ed. Richard Katz and Peter Mair, 24–29. Decline of the Working-Class Vote, the London: Sage. Reconfiguration of the Welfare Support Coalition and Kirchheimer, Otto. 1966. “The Transformation of the Consequences for the Welfare State.” Journal of Western European Party System.” In Political Parties European Social Policy 25(1): 50–75. and Political Development, ed. Joseph La Palombara and Goldberg, Andreas C. 2020. “The Evolution of Cleavage Myron Weiner, 177–200. Princeton, NJ: Princeton Voting in Four Western Countries: Structural, University Press. Behavioural or Political Dealignment?” European Kitschelt, Herbert. 1994. The Transformation of European Journal of Political Research 59(1): 68–90. Social Democracy. Cambridge: Cambridge University Goldthorpe, J. 1982. “On the Service Class, Its Formation Press. and Future.” In Social Class and the Division of Labour, Knutsen, Oddbjørn. 2007. “The Decline of Social Class?” ed. A. Giddens and G. Mackenzie, 162–185. In The Oxford Handbook of Political Behavior, Cambridge: Cambridge University Press. ed. Russell J. Dalton and Hans-Dieter Klingemann, Harbers, Imke. 2010. “Decentralization and the 457–480. Oxford: Oxford University Press. Development of Nationalized Party Systems in New ——. 2018. Social Structure, Value Orientations and Party Democracies.” Comparative Political Studies 43(5): Choice in Western Europe: A Comparative Longitudinal 606–27. Study. Basingstoke: Palgrave Macmillan. Hooghe, Liesbet, Gary Marks, and Carole J. Wilson. Kriesi, Hanspeter. 1998. “The Transformation of 2002. “Does Left/Right Structure Party Positions on Cleavage Politics: The 1997 Stein Rokkan Lecture.” European Integration?” Comparative Political Studies 35 European Journal of Political Research 33(2): 165–85. (8): 965–89. Lane, Jan-Erik, and Svante Ersson. 1987. Politics and Huntington, Samuel. 1991. The Third Wave: Society in Western Europe. London: Sage. Democratization in the Late Twentieth Century. Leiulfsrud, Håkon, Ivano Bison, and Heidi Jensberg. Norman: University of Oklahoma Press. 2005. Social Class in Europe: European Social Survey Imai, Kosuke, and In Song Kim. 2019. “When Should We 2002/3. Trondheim: NTNU Samfunnsforskning/ Use Unit Fixed Effects Regression Models for Causal NTNU Social Research Ltd. Inference with Longitudinal Data?” American Journal of Lijphart, Arendt. 1999. Patterns of Democracy: Political Science 63(2): 467–90. Government Forms and Performance in Thirty-six Inglehart, Ronald. 1977. The Silent Revolution: Changing Countries. New Haven, CT: Yale University Press. Values and Political Styles in Advanced Industrial Society. Lipset, Seymour Martin, and Stein Rokkan. 1967. Princeton, NJ: Princeton University Press. “Cleavage Structures, Party Systems and Voter ——. 1984. “The Changing Structure of Political Alignments: An Introduction.” In Party Systems and Cleavages in Western Societies.” In Electoral Change in Voter Alignments: Cross-National Perspectives, Advanced Industrial Democracies: Realignment or ed. Seymour Martin Lipset and Stein Rokkan, 1–64. Dealignment, ed. Russell Dalton, Scott C. Flanagan, New York: Free Press. and Paul Allen Beck, 25–69. Princeton, NJ: Princeton Mainwaring, Scott, and Edurne Zoco. 2007. “Political University Press. Sequences and the Stabilization of Interparty Jansen, Giedo, Geoffrey Evans, and Nan Dirk De Competition: Electoral Volatility in Old and New Graaf. 2013. “Class Voting and Left–Right Party Democracies.” Party Politics 13(2): 155–78. Positions: A Comparative Study of 15 Western Mainwaring, Scott, Carlos Gervasoni, and Annabella Democracies, 1960–2005.” Social Science Research España-Najera. 2017. “Extra-and Within-System 42(2): 376–400. Electoral Volatility.” Party Politics 23(6): 623–35. Johnson, Joel W., and Jessica S. Wallack. 2012. “Electoral Mair, Peter. 2007. “Left–Right Orientations.” In The Systems and the Personal Vote.” Accessed January Oxford Handbook of Political Behavior, ed. Russell J.

18 Perspectives on Politics